Filtern
Dokumenttyp
- Video (14)
- Konferenzveröffentlichung (5)
- Wissenschaftlicher Artikel (1)
- Masterarbeit (1)
- Arbeitspapier (1)
Sprache
- Englisch (22) (entfernen)
Volltext vorhanden
- ja (22) (entfernen)
Schlagworte
- Robotik (8)
- Flugkörper (7)
- UAV (7)
- Rettungsrobotik (5)
- 360° Panorama (1)
- Artificial Intelligence (1)
- Augmented Reality (1)
- Autonomous Agents (1)
- Codegenerierung (1)
- Enterprise JavaBeans (1)
- Erweiterte Realität <Informatik> (1)
- Hands-free Interaction (1)
- Human-Robot Interaction (1)
- Machine Learning (1)
- Multi-Agent System (1)
- NeRF (1)
- Object Recognition (1)
- Object-relational Mapping (1)
- Ortsbestimmung (1)
- Persistenz <Informatik> (1)
- Rescue Robotics (1)
- Small UAVs (1)
- Smart Grid (1)
- Visual Monocular SLAM (1)
- composition (1)
- design process (1)
- gender stereotypes (1)
- gender-sensitive design (1)
- gender-specific design (1)
Institut
- Informatik und Kommunikation (22) (entfernen)
Autonomy and self-determination are fundamental aspects of living in our society. Supporting people for whom this freedom is limited due to physical impairments is the fundamental goal of this thesis. Especially for people who are paralyzed, even working at a desk job is often not feasible. Therefore, in this thesis a prototype of a robot assembly workstation was constructed that utilizes a modern Augmented Reality (AR)-Head-Mounted Display (HMD) to control a robotic arm. Through the use of object pose recognition, the objects in the working environment are detected and this information is used to display different visual cues at the robotic arm or in its vicinity. Providing the users with additional depth information and helping them determine object relations, which are often not easily discernible from a fixed perspective.
To achieve this a hands-free AR-based robot-control scheme was developed, which uses speech and head-movement for interaction. Additionally, multiple advanced visual cues were designed that utilize object pose detection for spatial-visual support. The pose recognition system is adapted from state-of-the-art research in computer vision to allow the detection of arbitrary objects with no regard for texture or shape.
Two evaluations were performed, a small user study that excluded the object recognition, which confirms the general usability of the system and gives an impression on its performance. The participants were able to perform difficult pick and place tasks with a high success rate. Secondly, a technical evaluation of the object recognition system was conducted, which revealed an adequate prediction precision, but is too unreliable for real-world scenarios as the prediction quality is highly variable and depends on object orientations and occlusion.
An EJB container can host three types of beans: Session beans to model business processes, entity beans to represent business objects and message-driven beans to provide for asynchronous method calls. This paper addresses entity beans and their mapping to persistent storage, especially relational and object-relational databases. A tool named BeanMaker is presented which can do object mapping either automatically by metadata analysis of a database schema or manually based on intrinsic real world semantics supplied by the user. BeanMaker is a running prototype system with an intuitive GUI interface. This paper looks what's behind the scenes and focuses on design issues and concepts of code generation.
The article highlights gender codes in design, particularly in web design, by means of current examples. Different aspects of gender-specific design are looked at in detail and their inherent problems discussed: on the one hand the development of a special solution (gender-specific for women), on the other hand, web design with reduced functionality and simplification of information (i.e. image representation) which sometimes even leads to a negation of technology. The article illustrates that gender codes and stereotypical role models can be embodied on different design levels of web design (use and artefact): in structure/navigation, in creative elements by the use of shape, colour and imagery and on a textual level. These design decisions have an impact on the power of users to act, their individual gender identity and the structural gender identity/social perception of gender. The article demonstrates that gender codes in current web design are very present and aims to sensitize the topic.
Problem
- How to effectively use aerial robots to support rescue forces?
- How to achieve good flight characteristics and long flight times?
- How to enable simple and intuitive control?
- How to efficiently record image data of the environment?
- How to generate flight and image data for rescue forces?
Implementation:
The flying robot was designed in Autodesk Fusion360. In order to achieve high stability as well as low weight, the frame was milled from carbon. Mounts such as for GPS and 360° camera were 3D printed. A special feature is that the flying robot is not visible in the panoramic view of the 360° camera. The flight controller of the robot was set up using Ardupilot. The communication with the robot is done via MAVLink (UDP).To support different platforms, a software was realized as a web application. The front end was created using HTML, CSS and Javascript.
The back end is based on Flask-Socket-IO (Python). For the intelligent recognition of motor vehicles a micro controller with an integrated camera is used. For the post-processing of flight and video data a pipeline was implemented for automation.
The video shows a very high resolution 3D point cloud !!! of the outdoor area of the German Rescue Robotics Center. For the recording, a 25-second POI flight was performed with a Mavic 3. From the 4K video footage captured during this flight, 77 images were cropped and localized within 4 minutes using colmap and processed using Neural Radiance Fields (NeRF). The nerfacto model of Nerfstudio was trained on an Nvidia RTX 4090 for 8 minutes. In summary, a top 3D model is available to task forces after about 13 minutes. The calculation is performed locally on site by the RobLW of the DRZ. The video shown here shows a free camera path rendered at 60 hz (Full HD).
The two churches, San Francesco and Sant'Agostino in Amatrice, Italy was hit by an earthquake on August 24 2016. Both churches are in a state of partial collapse, in need of shoring to prevent potential further destruction and to preserve the national heritage. The video show the mission at 1.Sept.2016 in clips of 10 seconds.
The TRADR project was asked by the Italian firebrigade Vigili del Fuoco to provide 3D textured models of two churches.
The team entered San Francesco with two UGVs (ground robots) and one UAV (drone, flown by Prof. Surmann), teleoperating them entirely out of line of sight and partially in collaboration. We entered Sant'Agostino with one UAV (also flown by Prof. Surmann) while two other UAVs were providing a view from different angles to facilitate maneuvering them entirely out of line of sight.
Venice 2018: Tradr Review
(2018)
The video shows an orthopoto and a textured 3D model of the location. 300 images were recorded in two short flights with a Mavic Pro in 50 meter height. The first one was a single grid while the camera facing down and the second one was a double grid facing the camera at an 60 degree angle. The 3D model is computed with OpenDroneMap.
Challenging visual localization of an UAV while flying out of a room into a snowy environment (~ 4:50). The UAV is equipped with a 360° camera. The localization is done with OpenVSLAM.
The video was recorded in Jan. 2019 at the Fire Brigade training center in Dortmund
To achieve nearly real time conditions the original resolution of 5k (30 fps) was reduced to 2k (ffmpeg -i video.mp4 -vf scale=1920:-1 -crf 25 vido-small.mp4) with high compression (-crf 25). This reduce the original size from 3.2 GB to 93MB (~ 4 MBit/s which could be transmitted online via a radio link). The localization shown did not use frameskip. With a frameskip above 1 the localization fails while the UAV is flying through the window. Indoor localization can be done with a frameskip of 3 in real time.
360° Camera at a small UAV
(2021)
In the realm of digital situational awareness during disaster situations, accurate digital representations,
like 3D models, play an indispensable role. To ensure the
safety of rescue teams, robotic platforms are often deployed
to generate these models. In this paper, we introduce an
innovative approach that synergizes the capabilities of compact Unmaned Arial Vehicles (UAVs), smaller than 30 cm, equipped with 360° cameras and the advances of Neural Radiance Fields (NeRFs). A NeRF, a specialized neural network, can deduce a 3D representation of any scene using 2D images and then synthesize it from various angles upon request. This method is especially tailored for urban environments which have experienced significant destruction, where the structural integrity of buildings is compromised to the point of barring entry—commonly observed post-earthquakes and after severe fires. We have tested our approach through recent post-fire scenario, underlining the efficacy of NeRFs even in challenging outdoor environments characterized by water, snow, varying light conditions, and reflective surfaces.
This technical report is about the architecture and integration of very small commercial UAVs (< 40 cm diagonal) in indoor Search and Rescue missions. One UAV is manually controlled by only one single human operator delivering live video streams and image series for later 3D scene modelling and inspection. In order to assist the operator who has to simultaneously observe the environment and navigate through it we use multiple deep neural networks to provide guided autonomy, automatic object detection and classification and local 3D scene modelling. Our methods help to reduce the cognitive load of the operator. We describe a framework for quick integration of new methods from the field of Deep Learning, enabling for rapid evaluation in real scenarios, including the interaction of methods.
In this paper, we present a method for detecting objects of interest, including cars, humans, and fire, in aerial images captured by unmanned aerial vehicles (UAVs) usually during vegetation fires. To achieve this, we use artificial neural networks and create a dataset for supervised learning. We accomplish the assisted labeling of the dataset through the implementation of an object detection pipeline that combines classic image processing techniques with pretrained neural networks. In addition, we develop a data augmentation pipeline to augment the dataset with utomatically labeled images. Finally, we evaluate the performance of different neural networks.
This technical report is about the mission and the experience gained during the reconnaissance of an industrial hall with hazardous substances after a major fire in Berlin. During this operation, only UAVs and cameras were used to obtain information about the site and the building. First, a geo-referenced 3D model of the building was created in order to plan the entry into the hall. Subsequently, the UAVs were used to fly in the heavily damaged interior and take pictures from inside of the hall. A 360° camera mounted under the UAV was used to collect images of the surrounding area especially from sections that were difficult to fly into. Since the collected data set contained similar images as well as blurred images, it was cleaned from non-optimal images using visual SLAM, bundle adjustment and blur detection so that a 3D model and overviews could be calculated. It was shown that the emergency services were not able to extract the necessary information from the 3D model. Therefore, an interactive panorama viewer with links to other 360° images was implemented where the links to the other images depends on the semi dense point cloud and located camera positions of the visual SLAM algorithm so that the emergency forces could view the surroundings.
Renewable and sustainable energy production by many small and distributed producers is revolutionizing the energy landscape as we know it. Consumers produce energy, making them to prosumers in the smart grid. The interaction between prosumers and other entities in the grid and the optimal utilization of new smart grid components (electric cars, freezers, solar panels, etc.) are crucial for the success of the smart grid. The Power Trading Agent Competition is an open simulation platform that allows researchers to conduct low risk studies in this new energy market. In this work we present Maxon16, an autonomous energy broker and champion of the 2016's Power Trading Agent Competition. We present the strategies the broker used in the final round and evaluate the effectiveness of the strategies by analyzing the tournament's results.